Expansion of CD4+ CD25+ Foxp3+ T cells by bone marrow-derived dendritic cells

Authors

  • Ivo Marguti,

    1. Laboratory of Clinical Immunology, Department of Immunology, Instituto de Ciências Biomédicas, University of São Paulo, São Paulo, Brazil
    Search for more papers by this author
  • Guilherme Lopes Yamamoto,

    1. Laboratory of Clinical Immunology, Department of Immunology, Instituto de Ciências Biomédicas, University of São Paulo, São Paulo, Brazil
    Search for more papers by this author
  • Thaís Boccia da Costa,

    1. Laboratory of Clinical Immunology, Department of Immunology, Instituto de Ciências Biomédicas, University of São Paulo, São Paulo, Brazil
    Search for more papers by this author
  • Luiz Vicente Rizzo,

    1. Laboratory of Clinical Immunology, Department of Immunology, Instituto de Ciências Biomédicas, University of São Paulo, São Paulo, Brazil
    Search for more papers by this author
  • Luciana Vieira de Moraes

    1. Laboratory of Clinical Immunology, Department of Immunology, Instituto de Ciências Biomédicas, University of São Paulo, São Paulo, Brazil
    Search for more papers by this author

L. V. de Moraes, Departamento de Imunologia, Instituto de Ciências Biomédicas IV, Universidade de São Paulo. Av. Prof. Lineu Prestes, 1730, São Paulo 05508-000, Brazil. Email: luveira@usp.br
Senior author: Luciana Vieira de Moraes

Summary

Dendritic cells (DCs) are the most important antigen-presenting cells of the immune system and have a crucial role in T-lymphocyte activation and adaptive immunity initiation. However, DCs have also been implicated in maintaining immunological tolerance. In this study, we evaluated changes in the CD4+ CD25+ Foxp3+ T-cell population after co-culture of lymph node cells from BALB/c mice with syngeneic bone marrow-derived DCs. Our results showed an increase in CD4+ CD25+ Foxp3+ T cells after co-culture which occurred regardless of the activation state of DCs and the presence of allogeneic apoptotic cells; however, it was greater when DCs were immature and were pulsed with the alloantigen. Interestingly, syngeneic apoptotic thymocytes were not as efficient as allogeneic apoptotic cells in expanding the CD4+ CD25+ Foxp3+ T-cell population. In all experimental settings, DCs produced high amounts of transforming growth factor (TGF)-β. The presence of allogeneic apoptotic cells induced interleukin (IL)-2 production in immature and mature DC cultures. This cytokine was also detected in the supernatants under all experimental conditions and enhanced when immature DCs were pulsed with the alloantigen. CD4+ CD25+ Foxp3+ T-cell expansion during co-culture of lymph node cells with DCs strongly suggested that the presence of alloantigen enhanced the number of regulatory T cells (Tregs) in vitro. Our data also suggest a role for both TGF-β and IL-2 in the augmentation of the CD4+ CD25+ Foxp3+ population.

Introduction

Dendritic cells (DCs) are professional antigen-presenting cells with a unique capacity to activate T lymphocytes and induce adaptive immune responses. This capacity, however, is dependent on the activation state of the DCs. Upon encounter with pathogens and recognition of pathogen-associated molecular patterns (PAMPs), or in inflammatory environments, DCs become activated. The activation process involves several phenotypic and functional changes on DCs, including up-regulation of costimulatory molecules, migration to lymph nodes and production of cytokines, which, acting in concert, will lead to proper T-cell activation.1

In addition to their ability to initiate adaptive immune responses, DCs also play a role in generating T cells with regulatory properties and in maintaining peripheral tolerance. In the steady state, presentation of antigens derived from apoptotic bodies to T lymphocytes leads to tolerance induction.2 The mechanisms involved are not yet clear but there is published evidence showing that DCs that have internalized apoptotic bodies remain immature3–5 and therefore are not efficient at inducing effector T-cell responses.6–8 Several other mechanisms have also been implicated in the maintenance of peripheral tolerance. DCs treated with interleukin-10 (IL-10) were shown to induce anergy in CD4 and CD8 T cells.9,10 Further studies showed that IL-10-treated DCs11 or DCs generated from bone marrow precursors in the presence of this cytokine12 were capable of generating IL-10-producing T cells endowed with regulatory properties. Other studies have also highlighted the ability of DCs to delete T-cell clones in the periphery as a way of maintaining tolerance to self.13–15

Over the last decade, regulatory T cells (Tregs) have been described as an indispensable cell population in the maintenance of peripheral immunological tolerance.16 Tregs develop in the thymus17 and are characterized by the expression of CD4, CD25 and the transcription factor Foxp3.18 In the periphery, one of the requisites for survival of Tregs is their ability to respond to IL-2,19 and mice lacking this cytokine20 or components of its receptor21 develop spontaneous autoimmune syndrome. Interestingly, it has been demonstrated that Tregs can also be generated from naïve T cells in the periphery.22,23 Recently, the interactions between Tregs and DCs have been explored in an attempt to elucidate the contribution of these professional antigen-presenting cells to the maintenance of the Treg population in the periphery. Importantly, these studies have shown that Tregs can either proliferate or be generated from naïve T CD4+ lymphocytes after interaction with DCs.24–30

Data from our group showed that immature DCs pulsed with allogeneic apoptotic thymocytes were more efficient at expanding CD4+ CD25+ Foxp3+ cells in vitro compared with syngeneic apoptotic thymocytes. This observation raises questions regarding the contribution of some factors such as the nature of the antigen and the capacity to generate major histocompatibility complex (MHC)–peptide complexes to the generation of Tregs. Published data show that immature DCs that engulfed apoptotic allogeneic I-E+ B blasts were 1000 to 10 000 times more efficient at forming MHC–peptide complexes than preprocessed I-E peptide,31 which means that DCs have the capacity to efficiently present peptides extracted from phagocytosed cells. In the present study, we established an experimental system to elucidate the contribution of apoptotic allogeneic cells to the expansion of CD4+ CD25+ Foxp3+ cell population as well as the influence of the activation state of DCs in this process. We cultured mature or immature bone marrow-derived DCs, pulsed with apoptotic allogeneic thymocytes, with syngeneic lymph node cells. Our results show that mature and immature DCs are capable of expanding CD4+ CD25+ Foxp3+ T cells. Importantly, we observed an increased percentage of CD4+ CD25+ Foxp3+ cells in conditions where immature DCs were previously loaded with apoptotic allogeneic cells.

Materials and methods

Mice

Six- to 10-week-old BALB/c (H-2d) and C57Bl/6 (H-2b) female mice were obtained from the Department of Immunology animal facility at the University of São Paulo and kept in microisolator cages under specific pathogen-free conditions. Experiments were performed following the guidelines for animal use and care approved by the Ethics Committee on Animal Research at the Instituto de Ciências Biomédicas of the University of São Paulo.

Generation of immature DCs

Immature DCs were generated in vitro from bone marrow cells as described by Inaba et al.,32 with some modifications. In brief, bone marrow cells were removed from the femurs of BALB/c mice and cultured in complete medium [Dulbecco’s modified Eagle’s minimal essential medium (DMEM)] supplemented with 5% fetal bovine serum (Hyclone, Logan, UT), 10−5 m 2-mercaptoethanol (2-ME; Sigma Chemicals Co., St Louis, MO), 2 mm l-glutamine, 0·1 mm vitamins, 1 mm sodium pyruvate, 0·1 mm non-essential amino acids and 100 μg/ml gentamicine, all purchased from Gibco BRL (Rockville, NY), in a six-well plate (Sarstedt, Newton, NC) at 2 × 106 cells/ml in a total volume of 5 ml/well with 10 ng/ml of recombinant murine (rm) granulocyte–macrophage colony-stimulating factor (GM-CSF) for 7 days. rmGM-CSF was renewed on the 4th day of culture. For DC maturation, cells were stimulated for 24 hr with 1 μg/mL of lipopolysaccharide (LPS) on day 6.

Flow cytometry

Cells were incubated with anti-CD16/32 (Fc block) for 30 min at 4° in phosphate-buffered saline (PBS) containing 3% fetal calf serum (FCS) and 0·01% sodium azide [fluorescence-activated cell sorting (FACS) buffer]. Cells were labelled with fluorescent antibodies against CD11c conjugated to phycoerythrin (CD11c-PE), CD11b-fluorescein isothiocyanate (FITC), CD8-FITC, B220-FITC, MHC-II(I-Ad)-FITC, CD80-FITC, CD86-FITC, CD40-FITC, CD4-Cy and CD25-FITC (0·5 μg/106 cells; all purchased from Pharmingen, BD, San Diego, CA) and incubated for 30 min at 4°. Cells were washed with 1 ml of FACS buffer and re-suspended in 300 μl of the same buffer before analysis. For evaluation of Foxp3 expression, cells were labelled with anti-CD4 PE-Cy5 and anti-CD25-FITC antibodies and were fixed, permeabilized and labelled with anti-Foxp3-PE antibody according to the manufacturer’s instructions (PE anti-mouse/rat Foxp3 Staining Set; e-Biosciences, San Diego, CA). Cells were analysed by flow cytometry using a FACScalibur (Pharmingen, BD) with cellquest software (Pharmingen, BD).

Cytokine evaluation

Cytokine production by mature and immature DCs, pulsed or not with apoptotic cells, was evaluated using the commercial kit OptEIA (Pharmingen, BD) for IL-12, IL-10, tumour necrosis factor (TNF)-α and IL-2. For transforming growth factor (TGF)-β evaluation we used the TGF-β1 EMAX Immunoassay System (Promega, Madison, WI). After 6 days of culture, DCs were harvested from six-well plates and 5 × 105 cells/well were plated in 96-well plates. DCs were left immature or stimulated with LPS (1 μg/ml) for 24 hr. Then 2·5 × 106 apoptotic cells (1 : 5 DC to apoptotic cell ratio) were added or not to DC cultures. Supernatants were harvested 24 hr later and stored at −20° until use. Cytokine measurement was performed according to the manufacturer’s instructions.

Induction of apoptosis in thymocytes

Thymuses were obtained from 3-week-old C57Bl/6 female mice. A single-cell suspension was prepared and cells were plated in six-well plates at a concentration of 107 cells/ml. Cells were treated with 10−7 m dexamethasone and kept in a 37°, 5% CO2 incubator for 4 hr. Cells were washed with DMEM and re-suspended in complete medium for culture. For evaluation of apoptosis after treatment with dexamethasone, 106 cells were washed in HEPES buffer (10 mm HEPES, 150 mm NaCl, 5 mm KCl, 1 mm MgCl2 and 1·8 mm CaCl2), re-suspended in the same buffer (100 μl) and labelled with Annexin V-FITC for 20 min in the dark at room temperature. After incubation period, 350 μl of HEPES buffer was added to the suspension. Immediately before analysis, 40 μl of propidium iodide (100 μg/ml) was added. Cells were analysed by flow cytometry.

Phagocytosis assay

Immature and mature DCs (4 × 104) were incubated in 1·5-ml plastic conical tubes (Eppendorf, Hamburg, Germany) with varying quantities of apoptotic cells in order to obtain DC:apoptotic cell ratios of 1 : 1, 1 : 5 and 1 : 10. Cells were kept in contact for 2 and 4 hr at 37° and in a 5% CO2 atmosphere. After the incubation period, the DCs were transferred to glass slides by cytospin. The slides were stained using the Hema 3 Kit (Biochemical Sciences Inc., Swedesboro, NJ) and the number of DCs that had phagocytosed apoptotic cells was determined by optic microscopy.

DC-stimulated allogeneic T-cell culture

Immature or mature DCs (5 × 105) from BALB/c mice were co-cultured with splenocytes from C57Bl/6 mice at different responder:stimulator (1 : 5 and 1 : 10) cell ratios for 48, 72 or 96 hr in a 96-well U-bottom plate. [3H]Thymidine (1 μCi/well) was added for the last 18 hr. Plates were harvested and [3H]thymidine incorporation was evaluated in counts per minute (c.p.m.).

DCs and lymph node cell co-culture

Immature or mature DCs from BALB/c mice were cultured overnight in complete medium with apoptotic thymocytes from C57Bl/6 animals at a ratio of 1 : 5 (DCs:allogeneic cells). After the incubation period, allogeneic thymocyte debris was harvested from the culture by washing the cells with DMEM. DCs were co-cultured in complete medium with syngeneic lymph node cells at a ratio of 1 : 5 (DCs:lymph node cells) for 5 days. At the end of the co-culture period, lymph node cells were harvested with DMEM, washed once and characterized by flow cytometry.

Suppression assay

Spleens were harvested from BALB/c or C57Bl/6 mice and a single-cell suspension was prepared. After lysis of red cells with NH4Cl buffer, cells were washed with PBS, re-suspended in complete medium and tested for viability. Then 2·5 × 105 cells from BALB/c mice were co-cultured with 2·5 × 105 irradiated cells (3000 rads) from C57Bl/6 mice in a 96-well round-bottom plate together with different numbers of generated TCD25+ sorted cells using a flow cytometer cell sorter (FACSVantage SE; BD Biosciences, San Jose, CA) from immature DC cultures previously incubated with apoptotic allogeneic thymocytes in a total volume of 200 μl/well. After 5 days of culture, cells were pulsed with [3H]thymidine (1 μCi/well; Amersham International, Buckinghamshire, UK) for 18 hr. Cells were harvested and [3H]thymidine uptake was measured using a β-counter.

Depletion of CD4+ CD25+ T cells from lymph node cell suspensions

For the depletion of CD4+ CD25+ T cells from lymph node single-cell suspensions, we used the CD4+ CD25+ Regulatory T cell Isolation Kit (Miltenyi Biotec Inc., Auburn, CA). The depletion was performed following the manufacturer’s instructions. In all experiments, the purity obtained was > 97% in the population depleted of CD4+ CD25+ T cells.

Statistical analysis

Data are presented as mean values ± standard deviation. The Mann–Whitney test or analysis of variance (ANOVA) with Tukey’s post hoc test was performed using graphpad prism 4.0 software (GraphPad software Inc., La Jolla, CA). Data were considered significant at < 0·05.

Results

Characterization of bone marrow-derived DCs

DCs were generated in vitro from bone marrow cells of BALB/c mice using rmGM-CSF. At the end of the culture period, approximately 80% of these cells were CD11c+ CD11b+ CD4 CD8 B220 (data not shown). Upon stimulation with LPS, DCs increased expression of MHC-II and the costimulatory molecules CD80, CD86 and CD40 (Fig. 1). To evaluate the functional properties of bone marrow-derived DCs, we tested their capacity to induce the proliferation of allogeneic spleen cells. Immature or mature DCs were cultured with spleen cells from C57Bl/6 mice at different responder:stimulator cell ratios and the proliferative response was evaluated at different time-points. Splenocytes alone were used as a control. As shown in Fig. 2, both immature and mature DCs were capable of inducing allogeneic T-cell proliferation. Furthermore, at all time-points mature DCs were more efficient at stimulating proliferative responses than immature DCs.

Figure 1.

 Phenotypic analysis of immature dendritic cells (iDCs) and mature dendritic cells (mDCs). DCs were generated from bone marrow cells from BALB/c mice with recombinant murine granulocyte–macrophage colony-stimulating factor (rmGM-CSF). Cells were stimulated with lipopolysaccharide (LPS) or left unstimulated for 24 hr and were labelled with fluorescent antibodies against surface molecules. (inline image) isotype control; (inline image) surface marker. The median intensity of fluorescence (MIF) is indicated on the markers in the histograms. MHC, major histocompatibility complex.

Figure 2.

 Induction of allogeneic T-cell proliferation by bone marrow-derived dendritic cells (DCs). Immature or lipopolysaccharide (LPS)-matured DCs from BALB/c mice were co-cultured with total spleen cells from C57Bl/6 mice at different DC:T-cell ratios (1 : 5 and 1 : 10) for 48, 72 and 96 hr, in a 96-well U-bottom plate. Proliferation was measured by [3H]thymidine incorporation. Data are shown as mean ± standard deviation and are representative of three independent experiments. ***< 0·001 compared with immature DCs. c.p.m., counts per minute; iDC, immature DC; mDC, mature DC.

Phagocytosis of apoptotic thymocytes by DCs

Because we wanted to evaluate the role of apoptotic cells in the generation of CD4+ CD25+ Foxp3+ cells, we created conditions in which the greater part of the allogeneic cells would be apoptotic during the incubation period with DCs. Thymocytes from C57Bl/6 mice were incubated with dexamethasone for 4 hr, extensively washed and added to DC cultures for 18 hr. Phagocytosis of apoptotic thymocytes by immature and mature DCs at different DC:apoptotic cell ratios and at different time-points was evaluated. Data presented in Fig. 3 show the number of DCs that had phagocytosed apoptotic cells after 2 and 4 hr. No differences were observed between immature and mature DCs and there was a progressive and modest but not statistically significant increase in the phagocytosis rate when apoptotic cells were added in higher numbers.

Figure 3.

 Phagocytosis of apoptotic allogeneic thymocytes (APOs) by dendritic cells (DCs). Immature DCs (iDCs) or lipopolysaccharide (LPS)-matured DCs (mDCs) were incubated with APOs and phagocytosis was evaluated after 2 and 4 hr of incubation. Cells were collected and transferred to glass slides by cytospin and stained with haematoxylin and eosin. (a) An iDC (left); an apoptotic thymocyte (arrow) that has been phagocytosed by an iDC (right); (b) results are presented as the mean ± standard deviation of both time-points combined and are representative of three independent experiments.

Effect of apoptotic cells on DC maturation

We next evaluated the expression of MHC-II and costimulatory molecules, as well as the production of cytokines, by DCs after incubation with apoptotic allogeneic thymocytes. The mature DC population incubated with apoptotic allogeneic thymocytes showed a slight decrease in the percentage of MHC-IIhigh cells. No changes were observed in the expression of costimulatory molecules (CD80, CD86 and CD40) either in immature or in mature DCs after incubation with apoptotic cells (Fig. 4). The production of IL-10, IL-12, TNF-α (Fig. 5a) and TGF-β (Fig. 5b) by mature DCs was not altered in the presence of apoptotic cells. However, under these conditions, the secretion of IL-2 was increased (Fig. 5c). Regarding immature DCs, all cytokines with the exception of TGF-β were below the detection limit of the assay. Importantly, the presence of allogeneic apoptotic thymocytes induced only IL-2 production by DCs (Fig. 5c) but did not modulate the production of TGF-β (Fig. 5b). Taken together, these data show that, in our system, allogeneic apoptotic cells have no effect on the maturation state of DCs but are able to induce IL-2 production in immature DCs and enhance the secretion of this cytokine by mature cells.

Figure 4.

 Expression of major histocompatibility complex (MHC) class II and costimulatory molecules in immature dendritic cells (iDCs) or mature dendritic cells (mDCs) incubated or not with apoptotic allogeneic thymocytes (APOs). DCs were generated from bone marrow cells from BALB/c mice with recombinant murine granulocyte–macrophage colony-stimulating factor (rmGM-CSF), as described in the main text. On day 6, cells were stimulated with lipopolysaccharide (LPS; 1 μg/ml) or left unstimulated for 24 hr. On day 7, iDCs and mDCs were incubated with APOs for 18 hr. Cells were collected and evaluated for the expression of MHC-II, CD80, CD86 and CD40. Data are representative of three independent experiments. FITC, fluorescein isothiocyanate.

Figure 5.

 Cytokine analysis in the supernatants of mature dendritic cells (mDCs) and immature dendritic cells (iDCs) incubated or not with apoptotic allogeneic thymocytes (APOs). DCs were generated from bone marrow-derived cells from BALB/c mice with recombinant murine granulocyte–macrophage colony-stimulating factor (rmGM-CSF) for 6 days. On day 7, cells (5 × 105) were transferred to a 96-well flat-bottom plate and stimulated with lipopolysaccharide (LPS; 1 μg/ml) or left unstimulated in the presence or absence of APOs (2·5 × 106) for 24 hr. Supernatants were harvested and (a) interleukin (IL)-12, tumour necrosis factor (TNF)-α and IL-10, (b) transforming growth factor (TGF)-β and (c) IL-2 were measured by enzyme-linked immunosorbent assay (ELISA) in triplicate for each experimental condition. Results are presented as mean ± standard deviation and are representative of two independent experiments.

Evaluation of the CD4+ CD25+ Foxp3+ T-cell population after co-culture of lymph node cells with DCs

Our next step was to study the influence of the maturation state of DCs and the presence of apoptotic cells on the expansion of the CD4+ CD25+ Foxp3+ T-cell population in vitro. We cultured total lymph node cells from BALB/c mice with syngeneic DCs that were previously incubated with apoptotic allogeneic thymocytes from C57Bl/6 mice and then evaluated the percentage of CD4+ CD25+ Foxp3+ lymphocytes obtained after co-culture. Under all conditions, the CD4+ CD25+ Foxp3+ population was expanded compared with the population observed in the lymph nodes from naïve mice (Fig. 6a). The maturation state of DCs did not influence the expansion of Tregs, as both immature and mature DCs were capable of increasing the percentages of CD4+ CD25+ Foxp3+ cells (Fig. 6b). Interestingly, immature DCs pulsed with apoptotic allogeneic cells produced increased percentages of Tregs compared with other experimental conditions that were suppressive in vitro (Fig. 6b and d). Apoptotic syngeneic thymocytes were not as efficient as allogeneic thymocytes at inducing Tregs in immature DC cultures (Fig. 7a). We also evaluated IL-2 in the supernatants of DC–lymph node co-cultures in all conditions. Interestingly, IL-2 production was significantly enhanced in the supernatants of immature DC cultures that had previously been incubated with allogeneic apoptotic cells compared with other conditions (Fig. 6c) and compared with apoptotic syngeneic-incubated immature DC assays (Fig. 7b). This observation supports a direct correlation between production of IL-2 and expansion of CD4+ CD25+ Foxp3+ cells.

Figure 6.

 Generation of CD4+ CD25+ Foxp3+ cells after co-culture of lymph node cells (LNs) from naïve mice with dendritic cells (DCs) pulsed or not with apoptotic allogeneic thymocytes (APOs). Mature dendritic cells (mDCs) or immature dendritic cells (iDCs) were incubated with APOs at a ratio of 1 : 5 (DCs:APOs) for 18 hr. APOs were harvested and DCs were co-cultured with syngeneic total lymph node cells for 5 days (1 : 5 DC:lymph node cell ratio). Cells were collected and labelled with fluorescent antibodies for CD4, CD25 and Foxp3. (a) Dot plot and (b) bar chart presentations of CD4+ CD25+ Foxp3+ cells (gate on CD4+ cells) are shown. Lymph node cells alone are presented as a control. (c) Interleukin (IL)-2 was evaluated in the supernatants by enzyme-linked immunosorbent assay (ELISA). CD4+ CD25+ Foxp3+ cells generated in iDC cultures in the presence of APOs were sorted and added to a mixed lymphocyte reaction (MLR) assay (2·5 × 105 cells: 2·5 × 105 cells; responder cells : stimulator cells) at different ratios [T regulatory cells (Treg):T effector cells] in a 96-well U-bottom plate for 5 days. Proliferation was evaluated by [3H]thymidine incorporation. Data are presented as mean ± standard deviation. (d) Suppression assay. CD4+CD25+ cells generated in iDC cultures were sorted by FACS an added at different ratios (Treg: spleen cells) to an MLR (2.5 × 105: 2.5 × 105; responder cells: stimulator cells) in a 96-well U-bottom plate for 5 days. In (b) *P < 0.05 and **P < 0.01 (LN compared to all groups); in (c) **P < 0.01 and in (d) *< 0·05 and **< 0·01 compared with proliferation in the absence of Tregs. Results regarding the generation of CD4+ CD25+ Foxp3+ cells are representative of four independent experiments and those for the suppressor assay are representative of two independent experiments. c.p.m., counts per minute; FITC, fluorescein isothiocyanate; PE, phycoerythrin.

Figure 7.

 Syngeneic apoptotic thymocytes are less efficient at generating CD4+ CD25+ Foxp3+ cells compared with allogeneic apoptotic cells. Immature dendritic cells (iDCs) were incubated with allogeneic (ALOs) or syngeneic (SYNs) apoptotic thymocytes at a ratio of 1 : 5 (DCs:apoptotic cells) for 18 hr. Apoptotic cells were harvested and DCs were co-cultured with syngeneic lymph node cells for 5 days (1:5, DC:lymph node cell ratio). Cells were collected and labelled with fluorescent antibodies for CD4, CD25 and Foxp3. (a) Dot plot presentation of CD4+ CD25+ Foxp3+ cells (gated on CD4+ cells). Lymph node cells alone are presented as a control. (b) Interleukin (IL)-2 was evaluated in the supernatants of regulatory T cell (Treg) cultures by enzyme-linked immunosorbent assay (ELISA). Data are presented as mean ± standard deviation and are representative of two independent experiments; ***< 0·001.

Expansion of the CD4+ CD25+ Foxp3+ T-cell population in CD4+ CD25+-depleted cultures

The increase in the CD4+ CD25+ Foxp3+ cell population previously observed prompted us to investigate whether this was attributable to the generation of CD4+ CD25+ Foxp3+ from CD4+ CD25 T cells or to the expansion of the CD4+ CD25+ Foxp3+ population present in the lymph nodes of naïve animals. To address this question we co-cultured immature or mature DCs, pulsed or not with apoptotic allogeneic cells, with lymph node cells depleted of CD4+ CD25+ T cells. We were able to obtain a highly purified population, eliminating approximately 98% of the CD4+ CD25+ cells (Fig. 8a). Our results show an increase in the CD4+ CD25+ population when lymph node cells were co-cultured with immature and mature DCs previously pulsed or not with apoptotic allogeneic cells (Figs 8b). Regarding the Foxp3+ cells, in all groups these cells expanded, especially when immature DCs were previously pulsed with apoptotic thymocytes (Fig. 8b and c). It is worth noting that in all conditions analysed the increase in the CD4+ CD25+ Foxp3+ cell population observed in this setting was less pronounced than that observed when total lymph node cells were co-cultured with DCs (Fig. 7). These observations suggest that the increase in the CD4+ CD25+ Foxp3+ cell population in the former is attributable to the expansion of the 2% contaminating CD4+ CD25+ T cells and not to de novo generation of CD4+ CD25+ Foxp3+ T cells from naïve T cells. Interestingly, in all assays we also observed expansion of CD4+ CD25+ Foxp3 T cells (Fig. 8b and d). Taken together, these results suggest that in our model the increase in the CD4+ CD25+ Foxp3+ population is mainly attributable to the expansion of the CD4+ CD25+ Foxp3+ cells present in the lymph nodes. Furthermore, our data also emphasize that the CD4+ CD25+ Foxp3+ cells present in the lymph nodes are important for inhibition of overt naïve T-cell activation during the co-culture period.

Figure 8.

 Expansion of CD4+ D25+ Foxp3+ cells from lymph node cells initially depleted of CD4+ CD25+ cells, after co-culture with immature dendritic cells (iDCs) or lipopolysaccharide (LPS)-matured dendritic cells (mDCs) pulsed or not with allogeneic apoptotic thymocytes (APOs). mDCs or iDCs were incubated with APOs at a ratio of 1 : 5 (DCs:APOs) for 18 hr. APOs were harvested and DCs were co-cultured with syngeneic total lymph node cells (LNs) depleted of CD25+ cells for 5 days (1:5, DC:lymph node cell ratio). Cells were collected and labelled with fluorescent antibodies for CD4, CD25 and Foxp3. (a) Percentage of CD4+ CD25+ T cells in lymph node cell suspensions before (left) and after (right) depletion; (b) dot plot of CD25+ Foxp3+ cells after co-culture in different conditions (gated on CD4+ cells). (c) Percentage of CD4+ CD25+ Foxp3+ and (d) percentage of CD4+ CD25+ Foxp3 cells after co-culture with DCs. Data are representative of two independent experiments. FITC, fluorescein isothiocyanate; PE, phycoerythrin.

Discussion

In this study, we have shown that bone marrow-derived DCs from BALB/c mice are able to promote expansion of CD4+ CD25+ Foxp3+ T cells in vitro and that this occurs independently of the maturation state of the DCs as mature and immature cells are capable of expanding the Foxp3+ cell population. We have also shown that expansion of Tregs was favoured when immature DCs were incubated with allogeneic apoptotic thymocytes and that these cells were suppressive in vitro.

Our bone marrow-derived DCs showed phenotypic and functional characteristics of immature DCs. They expressed low levels of costimulatory molecules, were not able to produce inflammatory cytokines and had a reduced capacity to prime allogeneic T cells. Importantly, when activated with LPS they were very efficient at inducing T-cell proliferation, enhanced the expression of costimulatory molecules and were inflammatory cytokine producers. These results show that the DC population used throughout the study had features of immature DCs and that upon activation with LPS it were fully capable of maturing.

The efficiency of mature DCs in expanding CD4+ CD25+ Foxp3+ cells has previously been shown in mice25,33 and in humans.34 There is evidence in the literature that proliferation of Tregs induced by mature allogeneic or syngeneic DCs derived from bone marrow cells is dependent on exogenous IL-2.33 In our experiments, exogenous IL-2 was not necessary for expansion of CD4+ CD25+ Foxp3+ cells. We observed at least a twofold expansion of these cells when lymph node cells were incubated with mature or immature DCs. In our cultures, IL-2 production by DCs occurred only after maturation with LPS or incubation with apoptotic cells. The contribution of IL-2 secretion by DCs could explain in part Treg expansion. Importantly, in all co-culture conditions tested, IL-2 was detected in the supernatants, suggesting that lymph nodes cells also contribute to the secretion of this cytokine during co-culture. Although immature DCs were not IL-2 producers this cytokine was detected later in the supernatants of DC:lymph node cell co-cultures. We did not evaluate the source of IL-2 secretion in the co-cultures but we could speculate that allogeneic antigen could be activating naïve T cells thus inducing IL-2 production. This hypothesis is supported by our observation that in co-cultures where DCs were pulsed with apoptotic syngeneic cells lower IL-2 production was observed.

We loaded DCs with allogeneic apoptotic thymocytes and evaluated the influence of these DCs which had phagocytosed apoptotic bodies on CD4+ CD25+ Foxp3+ T-cell expansion. There are some points to be considered here. Because we wanted to evaluate the influence of exogenous antigens on Treg expansion and assure to a certain extent that immature DCs would remain immature during antigen presentation, we chose to work with allogeneic apoptotic cells. Many studies have shown that apoptotic cells can play an important role in the generation of peripheral tolerance, possibly by maintaining DCs in an immature state and/or inducing tolerogenic properties in these cells.35–37 In fact, our results showed that the maturation status of DCs was not altered when these cells were pulsed with allogeneic apoptotic thymocytes. For mature DCs, one possible explanation is that DCs were fully mature when apoptotic cells were added so the expected modulatory properties of the latter would not be capable of reversing such a state. Another point concerns MHC–peptide complex formation. It has been shown that DCs efficiently extract peptides from phagocytosed cells and they are therefore capable of forming more MHC–peptide complexes compared with DCs that were incubated with preprocessed peptides.31 In fact, our findings strongly suggest efficient alloantigen presentation as a result of increased percentages of CD4+ CD25+ Foxp3 cells in cultures where DCs were previously incubated with apoptotic cells. In these assays, lymph node cells were depleted of CD4+ CD25+ lymphocytes and expansion of CD4+ CD25+ Foxp3 cells was detectable in both immature and mature DC cultures. To a lesser extent, proliferation of the latter population was also observed in cultures that lacked allogeneic apoptotic thymocytes and was enhanced in mature DC assays. However, our data show that the number of Tregs generated after incubation of lymph node cells with syngeneic apoptotic thymocyte-pulsed DCs was decreased compared with cultures where DCs were previously incubated with allogeneic thymocytes. This event was also directly correlated to the production of IL-2 in DC:lymph node cell co-culture supernatants, suggesting that the nature of the antigen might be involved in the differential expansion of Tregs.

The experimental system in which lymph node cells were depleted of CD4+ CD25+ T lymphocytes could also contribute to knowledge of Treg expansion. In these conditions we were able to observe expansion of CD4+ Foxp3+ T cells in all groups. We could not rule out the possibility that the expansion of CD4+ CD25+ Foxp3+ lymphocytes observed, although depletion of CD4+ CD25+ cells was efficient (purity > 98%), could be attributable to the expansion of contaminating cells. Also population expands by conversion of a fraction of the CD4+ CD25 cells. However, as the expansion of the CD4+ CD25+ Foxp3+ T-cell population is much smaller in this experimental setting, the data suggest that in our system the CD4+ CD25+ Foxp3+ T-cell population increases mainly by expansion of this cell type, rather than de novo generation from CD4+ CD25 Foxp3 cells.

Studies have demonstrated that CD4+ CD25+ Foxp3+ cells are more efficiently expanded by mature bone marrow-derived DCs compared with immature DCs.25,26,38 Our data show that, despite the fact that both mature and immature DCs expand Tregs, presentation of allogeneic peptides in a non-inflammatory environment led to increased CD4+ CD25+ Foxp3+ cell expansion. The cytokine production profile of our generated DCs could support the differential Treg expansion observed in our study. We were unable to detect production of IL-12 and TNF-α by immature DCs, whereas in mature cells these cytokines were secreted at high levels. Both mature and immature DCs secreted significant amounts of TGF-β. This cytokine has been shown to play a role in the maintenance of the population, function and Foxp3+ expression of Tregs in the periphery.39,40 Furthermore, this cytokine has also been implicated in the generation of CD4+ CD25+ Foxp3+ cells from naïve T lymphocytes in different experimental conditions.22,28,29,40 It has recently been shown that differentiation of antigen-specific CD4+ CD25+ Foxp3+ cells from Foxp3 precursors and proliferation of these Tregs require DCs in the presence of TGF-β.41 Despite the enhanced production of pro-inflammatory cytokines by mature DCs, it is possible that the presence of TGF-β interferes with the activation of naïve T cells while helping with CD4+ CD25+ Foxp3+ maintenance and expansion in the cultures.

The expansion of Foxp3+ regulatory T cells has recently been shown to be favoured by suppressors of cytokine signalling (SOCS3)−/− DCs compared with wild-type DCs.42 These SOCS3−/− cells express lower levels of MHC class II molecules, CD40, CD86 and IL-12, a similar phenotype to that of the immature DCs in our study. In our system, the observation that both mature and immature DCs increased the percentages of CD4+ CD25+ Foxp3+ cells suggests that costimulatory molecules such as CD40 and CD80/86 might not be involved in this process.

Our system provides tools for elucidation of the mechanisms involved in DC and Treg interactions. The evaluation of the functional properties of our generated/expanded CD4+ CD25+ Foxp3+ cells in vivo is ongoing.

Acknowledgements

This work was supported by grants from Fundação de Amparo a Pesquisa do Estado de São Paulo (FAPESP) and the National Council for Scientific and Technologic Development (CNPq). LVdeM has a Post-Doctoral fellowship from FAPESP and LVR is a recipient of a personal award from CNPq. The authors wish to thank Christina Arslanian Kubo for technical assistance.

Ancillary